The Systemic Risks of OTC Derivatives
In a recent speech, Paul Tucker, Deputy Governor Financial Stability, Bank of England said :
“...it is an understatement that it would be a disaster if a clearing house failed. Commentators
have, indeed, been emphasising that CCPs are becoming systemic. To my own way of thinking,
they have already been systemic for the markets they clear for a very long time.”
This author agrees with Tucker. In this paper I set out some of the systemic risks which are arising as
the central clearing of OTC derivatives becomes more widespread. At the time of writing, this seems
inevitable: it has been mandated by the G-201, and various pieces of legislation such as the United
States’ Dodd Frank Act and the European Union’s proposed Regulation on OTC derivatives, central
counterparties and trade repositories require the clearing of certain OTC derivatives transactions.
Simultaneously the structure of clearing is changing, as central counterparties (‘CCPs’) are purchased
by or merge with larger corporations2. Therefore analysis of the systemic risks posed by the new
prominence of OTC derivatives central clearing is timely.
Systemic risk arises when an event, such as the distress of a financial institution or the failure of a
piece of market infrastructure, causes further failures or disruption. A CCP can be thought of as the
financial analogue of a dam, prevent rising water upstream from causing flooding downstream. One
problem with dams is that if they fail, those they are meant to protect can suffer significant distress.
Similarly CCPs can be both mitigate systemic risk and – if they become stressed – spread it. They can
also cause systemic risk through their policies, as we will demonstrate.
There are two principal interlinked causes of financial institution failure :
Solvency risk. Here losses cause the solvency of an institution to be doubtful.
Liquidity risk. Here an institution finds it difficult or expensive to meet claims when they
Both of these failure modes can affect clearing houses, and we discuss each of them in subsequent
sections. We also discuss some possible behavioural consequences of OTC derivatives clearing. The
paper concludes with some short suggestions for mitigating the risk of clearing house failure.
The G-20 Pittsburgh Communiqué states ‘All standardized OTC derivative contracts should be traded on
exchanges or electronic trading platforms, where appropriate, and cleared through central counterparties by
end-2012 at the latest’ *16+.
This process started with the demutualization of the CME in 2000. LSE’s takeover of LCH.Clearnet and the
proposed DB/NYSE merger are the latest examples of this trend towards CCPs as profit centres.
5 January 2012 1
In what follows we assume some familiarity with both central clearing and OTC derivatives, and
focus solely on systemic risk rather than issues of market efficiency. An introduction to central
clearing and further discussions of how to improve the bilateral and cleared derivatives markets can
be found in references [21,24,25,26,31].
1.1 Motivations for improving OTC derivatives market infrastructure
The financial crisis of 2007+ provided a number of motivations for improving financial market
infrastructure. These included:
The problem of interconnectedness. Contractual relationships between financial institutions,
including loans; lines of credit; securities financing transactions; and derivatives, give rise to
complex interdependencies between institutions. This can decrease financial stability in that
the failure of one or more actors causes losses at others in turn endangering them.
Transparency. In order to monitor the financial system, supervisors need to know who the
ultimate holders of risks are, and who is exposed to whom. If contractual relationships are
private then the financial system may be too opaque to be supervisable.
The OTC derivatives market was one example of these phenomena. It also suffered from some
specific issues which were, in retrospect, destabilizing including:
Some margin practices were highly procyclical. An example here is use of ‘trigger CSAs’,
whereby a counterparty had to post no margin initially but a potentially large amount if it
were to be downgraded below a certain threshold. These give rise to contingent liquidity
risk . Agreements of this type were a major factor in AIG’s difficulties.
Some market infrastructure was insufficiently robust. The period between a trade being
executed and all of the details of it being agreed between the two parties was variable and
sometimes extended. Furthermore, (mid market) valuations of all trades were not
necessarily agreed between counterparties, giving rise to uncertainty and the possibility of
1.2 A simple account of clearing
The central clearing of OTC derivatives introduces a central counterparty or CCP to sit between
cleared OTC derivatives market participants. Thus counterparty credit risk is concentrated at the
CCP, instead of being distributed amongst financial institutions. This simple view of the difference
between a bilateral and a cleared market is illustrated in Figures 1 and 2.
Clients Dealer 1 Dealer 2 Clients
Clients Dealer 3 Dealer 4 Clients
Figure 1: A simple view of a bilateral derivatives market
5 January 2012 2
Clients Dealer 1 Dealer 2 Clients
Clients Dealer 3 Dealer 4 Clients
Figure 2: A simple view of a central cleared derivatives market
1.3 Clearing complexities
The simple picture above is not accurate for various reasons. Principally:
CCPs do not (and should not) clear all OTC derivatives. Standardized products are often
liquid enough that CCPs can safely clear them: many non-standard products are not3.
Regulations require the use of multiple CCPs. Many jurisdictions require, or are likely to
require, OTC derivatives traded within their borders or referencing their currency, to be
cleared locally. This forces financial institutions to use multiple OTC derivatives CCPs.
Indeed, over 30 are currently either in operation or planned.
Dealer 1 Group Dealer 2 Group
Bilateral cpty Bilateral cpty
Clients CM for CCP1 CM for CCP1 Clients
CM for CCP1
CCP2 CM for CCP2
Dealer 3 Group Dealer 4 Group
Bilateral cpty Bilateral cpty
Clients CM for CCP1 etc CM for CCP1 Clients
CM for CCP1
CCP2 CM for CCP2
Figure 3: A somewhat more accurate picture of a central cleared market (with
bilateral relationships dashed and cleared ones solid for clarity)
There are different CCPs for different asset classes. Partly due to significant first mover
advantages, a variety of different CCPs have become important in different asset classes.
Some of these non-standard contracts may be capable of decomposition into standardized building blocks,
which would make them clearable. Others, including those designed to meet specific end-user risk
management needs, may not be amenable to this treatment. Current regulatory policy envisages these
uncleared contracts as attracting higher capital (and perhaps collateral) requirements. This could, of course,
affect the cost of some risk management products for end-users .
5 January 2012 3
Regulations sometimes require financial institutions to use different legal entities to ‘face’
CCPs. Thus for instance a firm must use a US domiciled vehicle known as an FCM (‘Futures
Commission Merchant’) to face a US CCP in the listed futures markets.
These factors complicate the picture, so that in fact a somewhat more accurate representation is
shown in Figure 34.
1.4 Infrastructure benefits that do not require clearing
It is worth noting that solving many of the issues identified in section 1.1 above does not require
central clearing. For instance, transparency can be addressed via the use of trade repositories, while
timely trade confirmation and ongoing reconciliation can be performed without the intercession of a
CCP. Moreover, if supervisors desire a change in margin practices , such as the abandonment of
trigger CSAs , then this could either be imposed by fiat, or encouraged via the regulatory capital
1.5 The key role of margin
We have already noted that CCPs concentrate counterparty credit risk. To protect themselves, they
have various defences, namely their financial resources. The first of these is margin. CCPs take both
initial and variation margin (‘IM’ and ‘VM’ respectively). The first of these protects the CCP against
the change in value of a portfolio of derivatives it has with a counterparty after default, whereas the
former reflects changes in value between trade date and the date of the call, as Figure 4 illustrates.
Portfolio value Close out
Default point point IM covers change in
value after default until
VM covers change in value up
to (close to) the default point
Figure 4: Initial and variation margin
An effective CCP will clear many OTC derivatives. Typically this will result in significant net positions
and hence calls for large amounts of margin. Thus for instance in 2010 LCH had cleared $248 trillion
notional of interest rate swaps, and had taken €58.5 billion of margin . With this much at stake,
the importance of margin practices cannot be over-estimated5. Who holds margin; how, when and
in what form it flows; and who can seize it under what conditions: these are important questions for
Cecchetti et al. *7+ assert that the “wider use of central counterparties (CCPs) for over-the-counter derivatives
has the potential to improve market resilience by lowering counterparty risk”. Whether this potential is
realized depends on the change in counterparty credit risk after clearing; and any increase in safety brought
about by the presence of CCP capital. Given the complexity of the figure, we conjecture that risk may well
increase; the second point though is harder to judge. See  for a further discussion of these issues.
Hand in hand with margin goes the netting of trades within a portfolio, because the margin on a given
cleared portfolio is called on the net rather than the gross risk of that portfolio. (There may in addition be
intra-client netting: this is a separate issue.) Given the ubiquity of portfolio netting, it is paramount that a
clearing house has the highest degree of certainty that its claim on a defaulting clearing member is net rather
5 January 2012 4
2. The systemic risks of clearing I: solvency
Before discussing how CCPs might fail, it is worth pointing out that some CCPs in the past have failed
, while others have required supervisory assistance . The question is therefore not entirely
academic, although it should be noted that historically CCP distress did not cause major financial
2.1 CCPs as risk concentrators
Supervisors are aware that successful OTC derivatives CCPs hold large amounts of counterparty
credit risk. Thus for instance Chairman of the Federal Reserve Board Ben Bernanke opined 
“increased reliance on clearinghouses to address problems in other parts of the system increases
further the need to ensure the safety of clearinghouses themselves.”
We begin by looking at the issues involved in making CCPs safe from solvency risk.
2.2 What are CCP financial resources for?
There are (at least) four functions that CCP financial resources must perform:
I. Reducing the loss given default (‘LGD’) to the CCP if a clearing member (‘CM’) fails;
II. Keeping market confidence in the CCP (and thus reducing its liquidity risk);
III. Absorbing losses; and
IV. Providing funds to continue to operate and recapitalize the CCP during and after resolution.
For current CCPs, I. Is typically met by initial margin and perhaps the defaulter’s default fund
contributions6; II. and III. by default fund and CCP equity, and IV. by a capital call mechanism.
2.3 CCP safety as a CDO pricing problem
A CCP is similar to a CDO in that:
It has a variety of assets, namely derivatives receivables from its counterparties;
It can suffer losses due to default;
It has a number of tranches of protection against the risk of default, namely the CCP’s
default waterfall . (In this analogy, margin is the junior tranche of the CDO, default fund
is a mezzanine tranche, and so on.)
CCPs only suffers losses once the lower tranches have been exhausted. Calculating the risk of this
occurring – and hence estimating the probability of a given reduction in CCP equity – is akin to
pricing a senior tranche in a CDO.
One key fact about pricing CDO senior tranches is the importance of default correlation7. If default
correlation is low, then senior tranches are typically safe even if defaults are somewhat more
frequent than expected: but if default correlation is higher than expected, then losses large enough
to damage senior tranches are much more probable. In the CCP context this means that we need
to understand what the default correlation of the CCP’s clearing members is likely to be in markets
Note that since we want to mitigate the consequences of the failure of each CM independently, LGD
mitigation has to be on a CM-by-CM basis, with no cross subsidy possible.
We use the term ‘default correlation’ loosely for the relationship between defaults. In fact this is often not
well captured by a single correlation; rather, copula approaches are required .
5 January 2012 5
stressed enough to cause their default. Unfortunately the fact pattern here is likely not comforting
for a CCP: markets will be volatile and illiquid, meaning that initial margin may well not cover close
out costs; and clearing members defaults will be more correlated, as they are all likely to suffer in
these conditions. This means that unless a CCP is safe against the default of a significant number of
clearing members in volatile markets, it could itself become distressed, possibly causing stress to
2.4 Operational risk
Another means by which a CCP might sustain losses which could threaten its solvency is operational
risk. For instance, were a CCP to discover that its margin model or netting documentation were not
legally enforceable in a major jurisdiction; or that its trade population was not balanced (i.e. there
were some cleared trades where there was a buyer but no seller or vice versa); or to suffer a major
rogue trader event in its treasury department, then the resulting loss could make the CCP insolvent,
or at least cause many clearing members to lose confidence in the CCP.
3. The systemic risks of clearing II: liquidity
The introduction of central clearing represents a profound change in the liquidity dynamics of the
financial system. Bilateral OTC agreements (especially interdealer agreements) typically do not
feature initial margin, and collateral can often be rehypothecated. In contrast, CCPs require initial
margin from both counterparties to a trade, and margin cannot typically be reused. The amounts at
stake for the financial system here are significant: $2-4 trillion by most informed estimates
[20,27,28,29]. Therefore the impact of central clearing on systemic liquidity cannot be ignored.
3.1 CCPs’ initial margin arrangements
CCPs calculate margin using proprietary models which estimate the amounts required to close out a
cleared portfolio. Different CCPs use different safety standards, data series, and calibration
methodologies in their analyses. Moreover CCPs typically have the right to change their margin
calculation at little notice, and to demand that margin based on their new calculation is posted on a
3.2 The liquidity implications of margin
Posting margin represents a real cost. This is moreover born by end users, in that they (or at least
their clearing members) have to post whatever margin is required. This is therefore one obvious
area of CCP competition: a CCP with lower margin requirements will, ceteris paribus, be more
attractive to some clients. However it will be less safe.
The effect of moving from a world with most collateral being rehypothecatable to one where much
of it cannot be reused is difficult to estimate8: certainly the liquidity available to many dealers will
fall markedly due to this change. Given the pressure from Basel III  to enhance liquidity
management – and the concomitant better pricing of liquidity – costs here could be significant. Of
course there are benefits too9, but it is surprising that there has been no comprehensive official
study of this phenomenon.
See references [18,28,29] for a further discussion.
For a discussion of some of the issues around the rehypothecation of collateral and the Lehman bankruptcy,
5 January 2012 6
3.3 Who owns margin: incentives
There are two principal CCP margin models10:
In the title transfer model, margin passes to the CCP. The CCP owns margin and can – absent
any other restrictions – do as it will with it. The margin poster has a claim on the CCP for the
return of its margin.
In the security interest model, the CCP has a claim over margin, but it does not legally own it.
Thus the poster receives any return on margin such as interest directly. A trustee is often
used in this situation to allow the poster to manage its total margin pool while ensuring that
sufficient amounts are available for the CCP.
The title transfer model poses the risk that the CCP is no longer just a clearer: it is an investor of
large amounts of margin. Returns on these investments accrue to the CCP, and thus it has an
incentive to manage them aggressively. There is clearly a risk that this will produce losses, which in
turn may threaten the solvency of the CCP11.
3.4 The procyclicality of margin
Margin requirements which rise in turbulent conditions draw liquidity from counterparties at the
times when they are likely to find it hardest to replace. Thus they are to be avoided . However,
CCPs which lowered initial margin requirements in good times to win business as discussed above, or
who simply followed the dictates of their risk sensitive models, will find themselves raising initial
margin in just such stressed markets12. In addition, some dealers will be facing variation margin
increases: one estimate here  is that this phenomenon alone could reduce the liquidity available
to the largest dealers by 28% over a week.
Procyclical margin practices do not endanger the CCP itself of course, but it could be destabilising for
its clearing members.
3.5 CCP runs
By analogy with a run on a bank, a run on a CCP occurs if one or more clearing members lose
confidence in the CCP, and begin to terminate their trades (or port positions to another CCP, should
this be possible13). One consequence of this is that the CCP has to return increasing amounts of
margin. If it cannot do this in a timely manner, perhaps because margin has been invested in
investments which prove illiquid in these conditions, then the CCP can suffer a liquidity crisis. It may
also suffer losses on rebalancing its portfolio. To avoid this, CCPs should keep sufficient financial
resources at all times to maintain confidence. Note that this continuity requirement means that
these resources are not available to absorb losses.
This discussion is necessarily brief and glosses over many legal complexities. For more details, see .
LCH Clearnet’s margin balances in 2010 was €58.5 billion *22+. Compare this with LCH’s equity of €319.9
million for the same year, and it can be seen that a loss of less than 1% of total margin can make a large CCP
insolvent. Now of course there are many controls in place to mitigate this risk, including the CCP’s governance,
and not all margin is owned by all CCPs, but nevertheless the equity/collateral leverage of many CCPs could
perhaps give systemic risk managers pause for thought.
Dramatic initial margin increases as markets deteriorate have already been seen in the context of repo
clearing of Irish and Portugese government bonds: see for instance
If CDS on CCPs were available, clearing members could protect themselves by buying protection in this way.
This approach is however problematic if such CDS must be cleared...
5 January 2012 7
4. Reducing the risks of CCPs
In this section we outline a dozen recommendations which aim to mitigate the risks identified above.
Many of these recommendations are based on the Basel Accords as it seems appropriate that
systemically important clearing houses should meet standards at least as constraining as those that
apply to internationally active banks14.
4.1 Financial resources requirements
A basic economic tenet of risk is that a party posing risk should pay the ‘fair price’ of that risk;
otherwise excessive risk taking is encouraged. In the context of clearing this means that a clearing
member’s contributions should be no less than the capital needed to support the counterparty
credit risk on their portfolio15. Therefore two basic regulatory recommendations to ensure the
safety of CCPs are –
RECOMMENDATION ONE (LGD Mitigation) All OTC derivatives CCPs should have systems in place to
allow them to calculate an estimate of the unexpected loss on any cleared portfolio under stressed
conditions to a comparable level of safety as that met by banks supervised under the Basel
Accords16. CCP margin should be arranged so that the initial margin required on any portfolio is
always greater than the unexpected loss estimate on that portfolio.
In what follows, we will denote by ‘excess DF’ the difference between the sum of all initial margins
and default fund contributions and the amount required to meet the LGD mitigation requirement
above for all cleared portfolios17. Excess DF is available to meet the functions (II.-IV.) identified in
RECOMMENDATION TWO (Estimating counterparty credit risk capital) All OTC derivatives CCPs
should have systems in place to allow them to calculate an estimate of stressed counterparty credit
risk capital required on their entire cleared portfolio using comparable models to those used by
banks supervised under the Basel Accords18.
The CPSS-IOSCO standards for CCPs [11,12] are less constraining than the current Basel rules. While CCPs
have different business models from banks, CCPs should only be permitted to meet lower standards than Basel
where this has been demonstrated to be prudent; thus far we are aware of no such demonstration. Of course,
some CCPs are banks and hence – presumably – meet Basel standards; but not all are.
The Basel Committee recognised this in their proposed criteria for deciding whether or not to risk weight
mutualised default fund contributions at penal levels . Notice here that some CCPs’ margin requirements
are not credit risk sensitive: here a lower credit quality CM pays the same margin and default fund as a higher
credit quality one despite being more likely to default. Some of these payments are LGD mitigation of course,
which is credit quality independent; but some of them supports counterparty credit risk, which isn’t.
Specifically, CCPs should calculate a stressed 99% ten day VAR on each cleared portfolio to Basel market risk
models standards . Ordinary VAR is highly procyclical and likely will not properly account for
autocorrelation/fat tails, hence a stressed VAR (i.e. a VAR based on stressed market data) is suggested.
Most clearing members will have at least two forms of IM: that for their own (house) portfolio; and that for
their clients (which is usually subject to some degree of segregation). Both house and client portfolios must be
separately mitigated in order to protect clients against the default of the CM. Therefore, IM must be adequate
to mitigate LGD for each class of portfolio separately.
Sophisticated CCPs should calculate CCR capital by (i) calculating stressed effective EPEs after applying LGD
mitigation using an IMM model; (ii) calculating EAD as 1.4 x stressed EPE; and (iii) combining these stressed
EADs to obtain a counterparty credit risk capital estimate at 99.9% confidence on a 1 year basis using a Basel
IRB approach with stressed default correlations . Note that stressed EPE will be zero for many portfolios if
IM is based on 10 day 99% stressed VAR; it is nevertheless important to perform the calculation.
5 January 2012 8
4.2 Financial resources requirements for operational risk
Any entity which turns over hundreds of trillions of dollars of transactions or which manages tens of
billions of dollars of investments has substantial operational risk. Much of a typical CCP’s financial
resources are not available to cover non-default losses: margin and default fund contributions, for
instance, can only be used to cover CCP losses due to counterparty non-performance. Typical
investment banks allocate several billion dollars of capital to operational risk alone19, yet many CCPs
have less $500M of equity capital. We therefore suggest the following –
RECOMMENDATION THREE (Estimating operational risk capital) All systemically important CCPs
should have systems in place to capture operational risk losses and to calculate operational risk
capital to a comparable level of safety as that required for sophisticated banks supervised under the
RECOMMENDATION FOUR (Minimum equity and alignment of interests) All systemically important
CCPs should have minimum common equity of $2B at all times21. A material fraction of CCP equity
should be at risk in the default waterfall before default fund contributions are exhausted.
4.3 Financial resources requirements for the two risks combined
One of the difficulties with a counterparty-by-counterparty assessment such as that suggested above
is that it does not account for multiple defaults. Clearly CCPs have some protection against the
failure of multiple clearing members, but a control is needed to ensure that it is enough, especially
due to the potential for clustered defaults with losses not completely absorbed by margin.
Therefore we suggest –
RECOMMENDATION FIVE (Minimum going concern financial resources) The sum of (a) excess DF;
and (b) CCP equity should be greater than the sum of: (1) the total capital required for the CCP’s
portfolio of counterparty credit risk; and (2) the capital required for operational risk.
4.4 Who owns margin: controls
Where CCPs own and manage margin for their own benefit, risk is clearly higher than if the CCP has
no incentive to extract extra return from collateral. At very least, clearing members should not be
forced to give title to their CCP . Therefore –
RECOMMENDATION SIX (Margin models) All systemically important CCPs which do not support the
security interest model for initial margin should make available third party custodial solutions under
which they have no financial interest in the return on margin or in its management. Clients of
clearing members should be able to opt for a third party custodian even if their clearing member
For instance, according to their annual report, JPMorgan’s economic capital estimate for operational risk
was $7.4B in 2010.
In other words, all systemically important CCPs should meet the criteria for AMA models under Basel 2 
and should calculate operational risk capital using such a model.
The FED views roughly 35 banks in the US as large enough to include in their Comprehensive Capital Analysis
and Review . All of these have more than $2B in capital. As Jean-Pierre Mustier of UniCredit put it, “The
‘thin capitalisation’ of CCPs – which will end up clearing hundreds of trillions of dollars notional in standardised
derivatives – could be compared with monoline insurance companies” . Therefore, even though $2B is
substantially higher than the capital of most current CCPs, we view it as a bare minimum for a systemically
important clearing house.
5 January 2012 9
Basel III  places significant constraints on bank liquidity risk. By analogy –
RECOMMENDATION SEVEN (Liquidity risk mitigation) All systemically important CCPs should have
a sufficient stock of high-quality liquid assets to cover estimated total net cash outflows over the
next 30 calendar days in highly stressed conditions.
4.5 The scope of clearing
CCPs need to be able to value all the products that they clear at all times. They need to be able to
risk manage them during the process of closing out a clearing member – something that will likely
occur during a stressed market. Furthermore there should be sufficient liquidity in all cleared
products for clearing members to be able to make accurate bids during the auction of a failed
clearing member’s portfolio. Otherwise, the failure of a clearing member may just push illiquid risk
back to other clearing members, potentially causing destabilising market moves as market
participants reduce risk. Finally, CCPs should not clear products which has substantial wrong way
risk, i.e. ones where changes in value of the product are likely to be highly correlated with the
default of clearing members. This suggests –
RECOMMENDATION EIGHT (Cleared product scope) CCPs should only clear OTC derivatives which
they can value and risk manage in stressed markets, and where the product is highly likely to be
liquid in the period after the default of one or more large clearing members. No product should be
cleared which has substantial wrong way risk.
4.6 CCP resolution
The G-20 sensibly recommended after the crisis that 
“Systemically important financial firms should develop internationally-consistent firm-specific
contingency and resolution plans”
In this context it is troubling that little progress has been made on an international framework for
CCP resolution. Given first that clearing houses have failed , and second that bank resolution has
been seen as sufficiently important it warranted its own G-20 commitment , it is surprising that
CCP resolution has received so little attention.
Following the aforementioned commitment, much work has however been done on bank
resolution22. Critical elements of a resolution regime for our purposes23 are:
“A process for early intervention with clear conditions governing their application;
Powers to operate and resolve the failing financial institution, including powers to
... continue needed contracts;
A mechanism to fund ongoing operations during the resolution process”
This suggests that supervisors should have some form of test for early intervention in a CCP. They
should have the power to take over an OTC derivatives CCP without such an event causing
See *15+ for the Financial Stability Board’s work. Various national initiatives, such as the UK’s Banking Act
2009 Special Resolution Regime are also worthy of note here.
 has a much more extensive discussion. The list given here is a subset of the elements recommended in
5 January 2012 10
termination of the CCP’s derivatives. Further, they require some mechanism to recapitalize the CCP
and fund its operations during and after resolution. This suggests –
RECOMMENDATION NINE (Intervention test) CCP supervisors should develop financial resources
and liquidity tests for early intervention in a CCP. The early intervention framework for systemically
important cross border CCPs should be internationally harmonized.
RECOMMENDATION TEN (CCP resolution) All OTC derivatives CCPs should have a living will which
sets out how they can be resolved. Their contractual documentation must permit resolution without
creating a termination event on cleared transactions. They must issue sufficient capital convertible
on resolution into equity24 to (a) fund the CCP’s operations during and 12 month period and (b)
absorb the maximum loss estimated to occur during a stress event25.
Combining the above, structure in Figure 5 is recommended for OTC derivatives CCPs’ capital
structure based on the suggested financial resources requirements26.
Bail-in capital To resolve CCP
Capital available to absorb losses
(Rest of) CCP equity
Remainder of default fund
(Fraction of) CCP equity
needed against risks run
Excess DF Defaulter’s default fund
Initial margin LGD mitigation
Figure 5: Recommended CCP capital structure and financial resources requirements
4.7 Who must clear
The vast majority of financial connections are between two different financial institutions rather
than between a financial institution and a non-financial27. Therefore much of the problem of
It is perhaps worth noting that having CCPs issue this kind of ‘bail in’ capital would mitigate the problem of
whether a CCP should or should not have access to the central bank window . This problem is complicated
in that some central banks, understandably, do not wish to be liquidity providers to the global derivatives
system in the event of crisis. However, without access to a central bank, a CCP could find itself unable to fund
itself as discussed above. The ability to convert existing instruments into substantial amounts of new equity
would mitigate this risk. (We prefer convertible instruments to capital calls as the capital is pre-funded in the
We are assuming that the stress event has caused the CCP to require intervention, and thus that it needs
this much additional capital in order to be capitally adequate. Of course, much more detail will need to be
supplied on how this stress loss is calculated and in particular on how many clearing members are assumed to
default during it.
See  for a related discussion of this and other resource structure including insurance. Op. cit. also has
the interesting suggestion that private contingent capital may have a role in meeting CCP financial resources
A simple if crude estimate of this for the OTC markets is the net to gross ratio for a given contract.
Examining the DTCC warehouse data for single name CDS, for instance, typical ratios for liquid sovereign CDS
are between 20 and 4, indicating that risk typically passes between at least three intermediaries between
ultimate buyer and ultimate seller.
5 January 2012 11
interconnectedness in the OTC derivatives markets is solved by requiring the clearing of interdealer
trades. Client clearing is problematic: very few clients can face CCPs directly; few clients are willing
and able to post margin in a form required by CCPs and to the required timeframe; and few clients
want to clear, presumably viewing the extra costs of clearing as unjustified given the scant benefit to
them. Moreover, forcing market participants to clear before CCPs are palpably safe means that
supervisors may be the cause rather than the cure to a subsequent crisis. Therefore we propose –
RECOMMENDATION ELEVEN (Mandatory clearing requirements) There should be no mandatory
clearing requirement until all systemically important clearing houses are in compliance with these
recommendations. There should be no requirement to clear OTC derivatives where one party is not
a financial institution.
Basel II rightly emphasised the importance of disclosure through the pillar 3 requirements . This
has a role to play in enhancing confidence in CCPs too, as disclosure from leading OTC derivatives
clearing houses is currently patchy. For instance, full details of margin and default fund calculations
are often not publically available, making it difficult to assess the safety of the CCP. Hence –
RECOMMENDATION TWELVE (Disclosure) All CCPs should be required to make timely public
disclosures to the Basel pillar 3 standard. All CCPs should disclose sufficient information to allow
potential clearing members or clients to make an informed decision as to their credit worthiness.
5. Behavioural effects
The financial system is not static. If change is imposed upon it, it will modify itself in reaction to that
change. Here some effects which might result from the imposition of OTC derivatives clearing are
briefly discussed. We look at the reactions of two classes of market participant: clearing members;
and their clients. These responses are necessarily difficult to predict (especially once higher order
reactions to reactions are considered), but that difficulty does not justify ignoring such effects.
5.1 Clearing member behaviour
OTC derivatives central clearing requires a considerable investment from clearing members and
significant ongoing costs. Moreover, the capital requirements for derivatives counterparty credit
risk are increasing generally; while the total capital required for cleared interdealer transactions is
lower on a single transaction basis than that required for the same transaction on a bilateral basis, it
by no means follows that central clearing has lower capital requirements than bilateral derivatives28.
Two changes will likely result from this:
Some banks will (perhaps partially) withdraw from the OTC derivatives markets, which in
turn may mean that bid/offer spreads are wider and risk management is more expensive;
Non-bank actors will likely play a larger part in the market. This will mean that bank
supervisors will have less understanding of underlying market flows and risk transfer than
This is partly due to the breaking of netting sets discussed in section 1.3, and partly due to increased capital
for default fund contributions .
5 January 2012 12
hitherto. A mechanism which increases the role of the non-bank financial system (aka ‘the
shadow banking system’) has consequences for financial stability.
5.2 Client behaviour
A rational corporate will look at a risk management problem by comparing the costs of hedging with
the cost of capital needed to support the risk. If the cost of hedging goes up, as it will with OTC
derivatives clearing, then more risk will be retained. Thus one effect of the imposition of clearing
will be that corporates hedge less risk. If this happens, the economic consequences could be
Mandatory central clearing of OTC derivatives has been introduced quickly and without a
comprehensive analysis of the risks it poses to the financial system. It represents a profound change
to an important market, and it will transform systemic risk dramatically. Some risks, such as
interdealer counterparty credit risk, may be reduced; but others will be increased. In this paper, a
number of potentially systemic risks which arise due to the risk of clearing have been identified.
Some tentative risk mitigation recommendations have also been made. Ultimately though this is
new territory: financial infrastructure is being changed by regulatory fiat, with results which will not
be wholly evident for some time29.
 Basel Committee on Banking Supervision, Basel II: International Convergence of Capital
Measurement and Capital Standards: A Revised Framework - Comprehensive Version, June 2006,
available at www.bis.org/publ/bcbs128.htm
 Basel Committee on Banking Supervision, Basel III: A global regulatory framework for more
resilient banks and banking systems - revised version, June 2011, available at
 Basel Committee on Banking Supervision, Consultative Document on Capitalisation of bank
exposures to central counterparties, 20th December 2010, available at
 Basel Committee on Banking Supervision, Report and Recommendations of the Cross-border
Bank Resolution Group, March 2010, available at www.bis.org/publ/bcbs162.htm
 B. Bernanke, Clearing and Settlement During the Crash, Review of Financial Studies, Volume 3, No
1, 1990, available at www.jscc.co.jp/en/ccp12/materials/docs/0416/5.pdf
 B. Bernanke, Clearinghouses, Financial Stability, and Financial Reform, Speech at 2011 Financial
Markets Conference, Stone Mountain, Georgia, April 4th 2011, available at
It is to be hoped, given this, that supervisors will review the impact of the OTC clearing via both quantitative
impact studies and via observing changes in market dynamics, and that policy changes will be made if the
evidence supports them.
5 January 2012 13
 S. Cecchetti, J. Gyntelberg, M. Hollanders, Central counterparties for over-the-counter derivatives,
BIS Quarterly Review, September 2009, available at www.bis.org/publ/qtrpdf/r_qt0909f.pdf
 U. Cherubini, E. Luciano, W. Vecchiato, Copula Methods in Finance, Wiley 2004
 J. Clark, The debate over CCP central bank liquidity, Risk, January 2011 available at
 Committee on the Global Financial System, The Role of Margin Payments and Haircuts in
Procyclicality, 2010, available at www.bis.org/publ/cgfs36.pdf
 CPSS-IOSCO, Guidance on the application of the 2004 CPSS-IOSCO Recommendations for Central
Counterparties to OTC derivatives CCPs, Consultative report, CPSS Publications No 89, May 2010,
available at www.bis.org/publ/cpss89.htm
 CPSS-IOSCO, Report on OTC derivatives data reporting and aggregation requirements -
consultative report, CPSS Publications No 96, August 2011, available at
 E. Davis, CCPs 'thinly capitalised' warns UniCredit's Mustier, Risk, 14th April 2011, available at
 Federal Reserve Board, Press Release, 10th June 2011, available at
 Financial Stability Board, Effective Resolution of Systemically Important Financial Institutions,
Consultative Document, 19th July 2011, available at
 G-20, Progress Report on the actions to promote financial regulatory reform, Pittsburgh G-20
Summit, 25th September 2009, available at
 A. Grody, Central counterparties — New uses for a century-old market mechanism, Journal of
Risk Management in Financial Institutions, Volume 4, No. 2, 2010.
 D. Heller, N. Vause, Expansion of Central Clearing, BIS Quarterly 2011 available at
 B. Hills, D. Rule, S. Parkinson, C. Young, Central Counterparty Clearing Houses and Financial
Stability, Bank of England Financial Stability Review 1999.
 ISDA, Margin Survey, 2011 available at www2.isda.org/functional-
 A. Lazarow, Lessons from International Central Counterparties: Benchmarking and Analysis,
Bank of Canada Discussion Paper 2011-4, available at www.bankofcanada.ca/wp-
5 January 2012 14
 LCH Clearnet, Annual Report 2010
 D. Murphy, Unravelling the causes of the credit crunch, Chapman Hall, 2009
 Report to the Supervisors of the Major OTC Derivatives Dealers on the Proposals of Centralized
CDS Clearing Solutions for the Segregation and Portability of Customer CDS Positions and Related
Margin, 2009, available at www.newyorkfed.org/markets/Full_Report.pdf
 P. Norman, The Risk Controllers: Central Counterparty Clearing in Globalised Financial Markets,
 C. Pirrong, The Economics of Central Clearing: Theory and Practice, ISDA Discussion Paper
Number One, May 2011
 E. Rowady, The Global Risk Transfer Market: Developments in OTC and Exchange-Traded
Derivatives, TABB Group 2010
 M. Singh, Collateral, Netting and Systemic Risk in the OTC Derivatives Market, IMF Working
paper 10/99, 2010, available at www.imf.org/external/pubs/ft/wp/2010/wp1099.pdf
 M. Singh, Making OTC Derivatives Safe―A Fresh Look, IMF Working paper 11/66, 2011,
available at www.imf.org/external/pubs/ft/wp/2011/wp1166.pdf
 M. Singh, J. Aitken, Deleveraging after Lehman—Evidence from Reduced Rehypothecation, IMF
Working paper 9/42, available at www.imf.org/external/pubs/ft/wp/2009/wp0942.pdf
 P. Tucker, Clearing houses as system risk managers, Speech at the launch of the DTCC-CSFI Post
Trade Fellowship Launch, 1st June 2011, available at
My thanks are due to the anonymous referees, Edwin Budding, Chris Holliman, Ulrich Karl and Craig
Pirrong for helpful comments on an earlier draft of this article. All errors remain, of course, my own.
5 January 2012 15